Two essays on utilitarianism, written from opposite points of view, by J. J. C. Smart and Bernard Williams. In the first part of the book Professor Smart advocates a modern and sophisticated version of classical utilitarianism; he tries to formulate a consistent and persuasive elaboration of the doctrine that the rightness and wrongness of actions is determined solely by their consequences, and in particular their consequences for the sum total of human happiness. This is a revised version of Professor Smart's famous essay “An outline of a system of utilitarian ethics’, first published in 1961 but long unobtainable. In part two Bernard Williams offers a sustained and vigorous critique of utilitarian assumptions, arguments and ideals, He finds inadequate the theory of action implied by utilitarianism, and he argues that utilitarianism fails to engage at a serious level with the real problems of moral and political philosophy, and fails to make sense of notions such as integrity, or even human happiness itself. Both authors are agreed on utilitarianism’s importance: it cuts across a number of different philosophical disputes and combines a systematic account of meta-ethical problems with a distinctive and substantive moral stand. It thus is, or involves, philosophy in both the traditional and the narrower, professional sense of the word, and is a key topic (often the first topic) in introductory philosophy courses. This book should also be of interest to welfare economists, political scientists and decision-theorists. CAMBRIDGE UNIVERSITY PRESS www.cambridge.org ISBN 0-521-09822-X UNI MM 098229)Copyrighted materialUTILITARIANISM for and against J.J.C. SMART Emeritus Professor, University of Adelaide Professor of Philosophy, Australian National University BERNARD WILLIAMS Provost of King's College, Cambridge 1] CAMBRIDGE UNIVERSITY PRESSPUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE The Pitt Building, Trumpington Street, Cambridge CB2 1RP, United Kingdom CAMBRIDGE UNIVERSITY PRESS The Edinburgh Building, Cambridge CB2 2RU, UK hep: //www.cup.cam.ac.uk 40) West 20th Sereet, New York, NY 10011-4211, USA http: //www.cup.org 10 Stamford Road, Oakleigh, Melbourne 3166, Australia © Cambridge University Press 1973 This book is in copyright. Subject to starmtory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1973 Reprinted 1975, 1976, 1978, 1980, 1982 (twice), 1983, 1985, 1988, 1989, 1990, 1991, 1995, 1996, 1997, 1998 Printed in the United States of America Typeset in Bembo Library of Congress Catalogue card number: 73-80487 ISBN 0-521-09822-X paperbackContents An outline of a system of utilitarian ethics J. J.C. SMART 1 Introductory 3 2 Act-utilitarianism and rule-utilitarianism 9 3__ Hedonistic and non-hedonistic utilitarianism 12 4 Average happiness versus total happiness 27 5 Negative utilitarianism. 28 6__Rightness and wrongness of actions 30 7_ The place of rules in act-utilitarianism 42 8 Simple application of game-theory technique 57 9 Utilitarianism and the future 62 10 Utilitarianism and justice 67 A critique of utilitarianism BERNARD WILLIAMS 1_Introductory 7 2_ The structure of consequentialism 82 3__ Negative responsibility: and two examples 93 4 Two kinds of remoter effect 100 5_Integrity 108 6 The indirect pursuit of utility 118 7_Social choice 135 Bibliography Ist J. J. C. SMARTCopyrighted materialAn outline of a system of utilitarian ethics Jj. J. C. SMARTCopyrighted material1, Introductory Such writers as J. §. Mill, H. Sidgwick and G, E, Moore, as a result of philosophical reflection, produced systems of normative ethics, Of recent years normative ethics has become distinguished from meta-ethics, which discusses the nature of ethical concepts. Indeed, as a result of the pre- valence of ‘non-cognitivist’ theories of meta-cthics, for example those of C. L. Stevenson! and R. M. Hare,? norma- tive ethics has fallen into some disrepute, at any rate as a philosophical discipline. For non-cognitivist theories of ethics imply that our ultimate ethical principles depend on our ultimate attitudes and preferences. Ultimate ethical princi- ples therefore seem to lie within the fields of personal deci- sion, persuasion, advice and propaganda, but not within the field of academic philosophy. While it is true that some ultimate ethical disagreements may depend simply on differences of ultimate preference, and while also the non-ultimate disagreements depend on differences about empirical facts, about which the philo- sopher is not specially qualified to judge, it nevertheless seems to me to be important to prevent this trend towards ethical neutrality of philosophy from going too far. The meta-ethical philosopher may far too readily forget that ordinary ethical thinking is frequently muddled, or else mixed up with questionable metaphysical assumptions. In the clear light of philosophical analysis some ethical systems may well come to seem less attractive. Moreover, even if there can be clear-headed disagreement about ultimate moral preferences, it is no small task to present one or other of the resulting ethical systems in a consistent and lucid manner, and in such a way as to show how 1 Ethies and Language (Yale University Press, New Haven, 1944). 2 The Language of Morals (Oxford University Press, London, 1952).4 An outline of a system of utilitarian ethics common, and often specious, objections to them can be avoided. It will be my object in the present study to state a system of ethics which is free from traditional and theological associations. This is that type of utilitarianism which R. B. Brandt has called ‘act-utilitarianism’.! Roughly speaking, act-utilitarianism is the view that the rightness or wrongness of an action depends only on the total goodness or badness of its consequences, i.c. on the effect of the action on the welfare of all human beings (or perhaps all sentient beings). The best sustained exposition of act-utilitarianism is, I think, that in Sidgwick’s Methods of Ethics,? but Sidgwick stated it within the framework of a cognitivist meta-cthics which supposed that the ultimate act-utilitarian principles could be known to be true by some sort of intellectual intuition. I reject Sidgwick’s meta-ethics for familiar reasons, and for the purpose of this study will assume the truth of some such ‘non-cognitivist’ meta-ethical analysis as that of Hare’s Language of Morals, or possibly that of D, H. Monro in his Empiricism and Ethics’ (Monro’s theory should perhaps be classed as subjectivist rather than as non-cognitivist. How- ever I am inclined to think that in the present state of linguistic theory it is not possible to make a very sharp distinction between these two sorts of theory.4 For our present purposes the distinction is unimportant, because both sorts of theory imply that a man’s ultimate ethical principles depend on his attitudes or feelings.) In adopting such a meta-cthics, I do, of course, renounce the attempt to 1 See R. B. Brandt, Evhical Theory (Prentice-Hall, Englewood Cliffs, New Jersey, 1959), p. 380. Brandt distinguishes ‘act’ utilitarianism from ‘rule’ utilitarianism. 7H. Sidgwick, Methods of Ethics, 7th ed. (Macmillan, London, 1962). 2D. H. Monro, Empiricism and Bthics (Cambridge University Press, London, 1967). *See my review of Monro's book, Philosophical Review 78 (1969) 259-61.1 Introductory 5 prove the act-utilitarian system. 1 shall be concerned with stating it in a form in which it may appear persuasive to some people, and to show how it may be defended against many of the objections which are frequently brought up against utilitarianism, Nevertheless I should like to indicate my opinion that the choice of conceptually clear and emotionally attractive systems of normative ethics which might be alternatives to it is not as wide as is sometimes thought. In the first place, B. H. Medlin’ has argued that it is impossible to state ethical egoism without cither confusion or else a sort of pragmatic inconsistency. Secondly, some widespread ethical systems depend partly on metaphysical premisses, and can therefore be undermined by philosophical criticism of these metaphysical bases. | myself would be pre- pared to argue that this is the case with respect to so-called ‘natural law’ ethics, which depends on a quasi-Aristotelian metaphysics. Thirdly, any system of deontological ethics, that is any system which does not appeal to the consequences of our actions, but which appeals to conformity with certain rules of duty, is open to a persuasive type of objection which may well be found convincing by some of those people who have the welfare of humanity at heart. For though, conceiv- ably, in most cases the dictates of a deontological ethics might coincide with those of human welfare and of an act- utilitarian ethics, there must be some possible cases in which the dictates of the system clash with those of human welfare, indeed in which the deontological principles prescribe actions which lead to avoidable human misery. In the most attrac- tive forms of deontological ethics the conflict with utili- tarianism is in consequence of some principle of ‘justice’ or ‘fairness’, and I shall revert to this issue later? In other cases, ‘Ultimate principles and ethical egoism’, Australasian Journal of Phil- osophy 35 (1957) 111-18, ? See pp. 67-73 below.6 An outline of a system of utilitarian ethics however, the conflict can be traced to some sort of confusion, perhaps even to some sort of superstitious ‘rule worship’, There is prima facie a necessity for the deontologist to defend himself against the charge of heartlessness, in his apparently preferring abstract conformity to a rule to the prevention of avoidable human suffering, Of course some deontologists might claim that though it is logically possible that their principles might conflict with the utilitarian one, in fact such a conflict would never occur. It seems that if such a deontol- ogy did exist, the utilitarian need not be concerned to defend himself against it, since its practical consequences would not differ from those of utilitarianism. However all deontological systems which are known to me do seem to differ from utilitarianism not only in theory but also in practice. Such a ‘persuasive’ objection to deontology is possible simply because we have assumed the truth of non-cognitivist (or possibly, subjectivist) meta-cthics, A cognitivist in meta- ethics of the type of Sir David Ross! could resist any such appeal to the heart by saying that whether we like it or not his deontological principles can be seen to be true, That they might sometimes conflict with human happiness or welfare might seem to him to be more of sentimental than of philo- sophic concern, But if we strip off the cognitivist meta- ethics from Ross's theory, then his deontology may come to look artificial and perhaps infected by a sort of ‘rule worship’. For example the obligation to keep promises seems to be too artificial, to smack too much of human social conventions, to do duty as an ultimate principle, On the other hand itis, as we shall see, harder to produce persuasive arguments against a restrained deontology which supple- ments the utilitarian principle by principles related to abstract justice and fair distribution. However, I am not attempting a eo Foundations of Ethies (Oxford University Press, London, 1939,1. Introductory 7 to show that the utilitarian can have no philosophically clear- headed rivals, but am merely trying to suggest that it is harder than is commonly believed to produce clear-headed and acceptable deontological systems of ethics, and that the range of these is probably not so wide as to embrace some of the well-known ones, such as that of Sir David Ross. In setting up a system of normative ethics the utilitarian must appeal to some ultimate attitudes which he holds in common with those people to whom he is addressing him- self, The sentiment to which he appeals is generalized benevolence, that is, the disposition to seek happiness, or at any rate, in some sense or other, good consequences, for all mankind, or perhaps for all sentient beings. His audience may not initially be in agreement with the utili- tarian position, For example, they may have a propensity to obey the rules of some traditional moral system into which they have been indoctrinated in youth. Nevertheless the utilitarian will have some hope of persuading the audi- ence to agree with his system of normative ethics, As a utilitarian he can appeal to the sentiment of generalized benevolence, which is surely present in any group with whom it is profitable to discuss ethical questions. He may be able to convince some people that their previous disposi- tion to accept non-utilitarian principles was due to concep~ tual confusions. He will not be able to convince everybody, no doubt, but that utilitarianism will not be accepted by everybody, or even by all philosophically clear-headed people, is not in itself an objection to it. It may well be that there is no ethical system which appeals to all people, or even to the same person in different moods, I shall revert to this matter later on. To some extent then, I shall be trying to present Sidgwick in a modern dress. The axioms of utilitarianism are no longer the deliverances of intellectual intuition but the expressions 1 See pp. 72-3 below.8 An outline of a system of utilitarian ethics of our ultimate attitudes or feelings. Deductions from these axioms nevertheless go through in very much the same way. In a discussion note commenting on the earlier edition of this monograph, Charles Landesman suggested' that as a non-cognitivist I am not entitled to talk about the logical consequences of ethical principles. However it is not clear to me that this is an insuperable difficulty. For example, R. M. Hare? and others have worked out theories of logical relations between imperative sentences, and even mere ex- pressions of attitude can be said to be consistent or inconsis- tent with one another. Thus “Boo to snakes’ is consistent with “Boo to reptiles’ and inconsistent with “Hurrah for reptiles’. Indeed there is no reason why a non-cognitivist should refuse to call ethical sentences ‘true’ or ‘false’. He can say ‘“Smith is good” is true if and only if Smith is good,’ He can even say things like ‘Some of Buddha’s ethical sayings are true’, thus giving to understand that he would be in agreement with some of the attitudes expressed in Buddha’s sayings, even though he is not telling, and even may not know, which ones these are. I must concede, however, that there are difficulties (attested to by the word ‘would’ in the previous sentence) in giving a proper semantics on these lines. ‘The semantics for ‘would’ gets us into talk about possible worlds, which are dubious entities. Again consider a sentence like ‘If it rains Smith’s action is right.’ A non-cognitivist would perhaps interpret this as expressing approval of Smith’s action in a possible world in which it is raining. However ethics, whether non- cognitivist or not, probably needs the notion of a possible world,? dubious or not, since it is concerned with alternative possible actions, and so in this respect the non-cognitivist '*A note on act utilitarianism’, Philosophical Review 73 (1964) 243-7. 2 The Language of Morals. 4 See R. Montague, ‘Logical necessity, physical necessity, ethies, and quantifiers’, Inquiry 3 (1960) 259-69.2. Act-utilitarianise and rule-utilitarianism 9 may not really be worse off than the cognitivist. At any rate, I am assuming in this monograph that adequate non-cogni- tivist theories of meta-ethics exist. 2. Act-utilitarianism and rule-utilitarianism The system of normative ethics which I am here concerned to defend is, as I have said earlier, act-utilitarianism. Act- utilitarianism is to be contrasted with rule-utilitarianism. Act-utilitarianism is the view that the rightness or wrong- ness of an action is to be judged by the consequences, good or bad, of the action itself. Rule-utilitarianism is the view that the rightness or wrongness of an action is to be judged by the goodness and badness of the consequences of a rule that everyone should perform the action in like circum- stances. There are two sub-varieties of rule-utilitarianism according to whether one construes ‘rule’ here as ‘actual rule’ or ‘possible rule’. With the former, one gets a view like that of §. E. Toulmin' and with the latter, one like Kant’s.? That is, if it is permissible to interpret Kant’s prin- ciple “Act only on that maxim through which you can at the same time will that it should become a universal law’ as “Act only on that maxim which you as a humane and benevolent person would like to see established as a uni- versal law.’ Of course Kant would resist this appeal to human feeling, but it seems necessary in order to interpret his doctrine in a plausible way. A subtle version of the Kantian type of rule-utilitarianism is given by R. F. Harrod in his ‘Utilitarianism Revised’? + An Examination of the Place of Reason in Ethics (Cambridge University Press, London, 1950). ?Immanuel Kant, Groundwork of the Metaphysic of Morals. Translated from the German in The Moral Law, by H. J. Paton (Hutchinson, London, 1943). 3 Mind 45 (1936) 137-56.10 An outline of a system of utilitarian ethics [have argued elsewhere! the objections to rule-utilitarian- ism as compared with act-utilitarianism.? Briefly they boil down to the accusation of rule worship: the rule-utilitarian presumably advocates his principle because he is ultimately concerned with human happiness: why then should he advo- cate abiding by a rule when he knows that it will not in the present case be most beneficial to abide by it? The reply that in most cases it is most beneficial to abide by the rule seems irrelevant. And so is the reply that it would be better that everybody should abide by the rule than that nobody should. This is to suppose that the only alternative to ‘everybody does A’ is ‘no one does A’, But clearly we have the possi- bility ‘some people do A and some don’t’, Hence to refuse to break a generally beneficial rule in those cases in which it is not most beneficial to obey it seems irrational and to be a case of rule worship, The type of utilitarianism which I shall advocate will, then, be act-utilitarianism, not rule-utilitarianism. David Lyons has recently argued that rule-utilitarianism (by which, I think, he means the sort of rule-utilitarianism which I have called the Kantian one) collapses into act- utilitarianism,.* His reasons are briefly as follows. Suppose ‘In my article “Extreme and restricted utilitarianism’, Philosophical Quarterly 6 (1936) 344-54. This contains bad errors and a better version of the article will be found in Philippa Foot (ed.), Theories of Ethics (Oxford University Press, London, 1967), or Michael D, Bayles (ed.), Contemporary Utilitarianism (Doubleday, New York, 1968). In this article [ used the terms ‘extreme’ and ‘restricted’ instead of Brandt's more felicitous ‘act’ and ‘rule’ which I now prefer. ? For another discussion of what in effect is the same problem see A. K. Stout's excellent paper, ‘But suppose everyone did the same’, Australa- sian Journal of Philosophy 32 (1954) 1-29. * On rule worship see I. M. Crombie, ‘Social clockwork and utilitarian morality’, in D, M, Mackifinon (ed.), Christian Faith and Communist Faith (Macmillan, London, 1953). See p. 109. 4 David Lyons, The Forms and Limits of Utilitarianism (Oxford University Press, London, 1965). Rather similar considerations have been put2. Act-utilitarianis and rule-utilitarianism Ir that an exception to a rule R produces the best possible consequences. Then this is evidence that the rule R should be modified so as to allow this exception. Thus we get a new rule of the form ‘do R except in circumstances of the sort C’, That is, whatever would lead the act-utilitarian to break a rule would lead the Kantian rule-utilitarian to modify the rule. Thus an adequate rule-utilitarianism would be extensionally equivalent to act-utilitarianism, Lyons is particularly interested in what he calls ‘threshold effects’, A difficulty for rule-utilitarianism has often appeared to be that of rules like ‘do not walk on the grass’ or ‘do not fail to vote at an election’. In these cases it would seem that it is beneficial if some people, though not too many, break the rule. Lyons points out that we can distinguish the action of doing something (say, walking on the grass) after some largish number n other people have done it from the action of doing it when few or no people have done it.When these extra circumstances are written into the rule, Lyons holds that the rule will come to enjoin the same actions as would the act-utilitarian principle. However there seems to be one interesting sort of case which requires slightly different treatment. This is the sort of case in which not too many people must do action X, but each person must plan his action in ignorance of what the other person does. That is, what A does depends on what B does, and what B does depends on what A does. Situations possessing this sort of circularity will be discussed below, pp. 57-62. Iam inclined to think that an adequate rule-utilitarianism would not only be extensionally equivalent to the act- utilitarian principle (ic. would enjoin the same set of actions forward by R. M. Hare, Freedom and Reason (Oxford University Press, London, 1963), pp. 131-6, and R, B, Brandt, “Toward a credible form of utilitarianism’, in H. N. Castafieda and G. Nakhnikian, Morality and the Language of Conduct (Wayne State University Press, Detroit, 1963), esp. pp. 119-23.12 An outline of a systent of utilitarian ethics as it) but would in fact consist of one rule only, the act- utilitarian one: ‘maximize probable benefit’. This is because any rule which can be formulated must be able to deal with an indefinite number of unforeseen types of contingency. No rule, short of the act-utilitarian one, can therefore be safely regarded as extensionally equivalent to the act- utilitarian principle unless it is that very principle itself. I therefore suggest that Lyons’ type of consideration can be taken even further, and chat rule-utilicarianism of the Kan- tian sort must collapse into act-utilitarianism in an even stronger way: it must become a ‘one-rule’ rule-utilitarianism which is identical to act-utilitarianism. In any case, whether this is correct or not, it is with the defence of act-utilitarian- ism, and not with rule-utilitarianism (supposing that there are viable forms of rule-utilitarianism which may be dis- tinguished from act-utilitarianism) that this monograph is concerned, (Lyons himself rejects utilitarianism.) 3- Hedonistic and non-hedonistic utilitarianism An act-utilitarian judges the rightness or wrongness of actions by the goodness and badness of their consequences. But is he to judge the goodness and badness of the conse- quences of an action solely by their pleasantness and un- pleasantness? Bentham,’ who thought that quantity of pleasure being equal, the experience of playing pushpin was as good as that of reading poetry, could be classified as a hedonistic act-utilitarian, Moore,? who believed that some ' Jeremy Bentham’s most important ethical work is “An Introduction to the Principles of Morals and Legislation’, in A Fragment on Government and an Introduction to the Principles of Morals and Legislation, ed. Wilfrid Harrison (Blackwell, Oxford, 1948), For the remark on poetry and pushpin sce Bentham’s Works (Tait, Edinburgh, 1843), vol. 2, pp. 3534 ; é. E Moore, Principia Ethica (Cambridge University Press, London, 1962).3. Hedonistic and non-hedonistic utilitarianism 3 states of mind, such as those of acquiring knowledge, had intrinsic value quite independent of their pleasantness, can be called an ideal utilitarian. Mill seemed to occupy an inter- mediate position.? He held that there are higher and lower pleasures. This seems to imply that pleasure is a necessary condition for goodness but that goodness depends on other qualities of experience than pleasanmess and unpleasantness. I propose to call Mill a quasi~ideal utilitarian, For Mill, pleasantness functions like x in the algebraic product, x x x 2. Ifx = o the product is zero. For Moore pleasantness functions more like x in (x +1) x y x z. If x= the product need not be zero. Of course this is only a very rough analogy. What Bentham, Mill and Moore are all agreed on is that the rightness of an action is to be judged solely by conse- quences, states of affairs brought about by the action, Of course we shall have to be careful here not to construe ‘state of affairs’ so widely that any ethical doctrine becomes utili- tarian. For if we did so we would not be saying anything at all in advocating utilitarianism. If, for example, we allowed ‘the state of having just kept a promise’, then a deontologist who said we should keep promises simply because they are promises would be a utilitarian, And we do not wish to allow this. According to the type of non-cognitivist (or subjectivist) ethics that I am assuming, the function of the words ‘ought’ and ‘good’ is primarily to express approval, or in other words, to commend. With ‘ought’ we commend actions. With ‘good’ we may commend all sorts of things, but here I am concerned with ‘good’ as used to commend states of affairs or consequences of actions. Suppose we could know with certainty the total consequences of two alternative actions A and B, and suppose that A and B are the only possible actions open to us. Then in deciding whether we 3], S. Mill, Utilitarianism, ed. Mary Warnock (Collins, London, 1962).14 An outline of a system of utilitarian ethics ought to do A or B, the act-utilitarian would ask whether the total consequences of A are better than those of B, or vice versa, or whether the total consequences are equal. That is, he commends A rather than B if he thinks that the total consequences of A are better than those of B. But to say ‘better’ is itself to commend. So the act-utilitarian has to do a double evaluation or piece of commending. First of all he has to evaluate consequences. Then on the basis of his evaluation of consequences he has to evaluate the actions A and B which would lead to these two sets of consequences. It is easy to fail to notice that this second evaluation is needed, but we can see that it is necessary if we remind ourselves of the following fact. This is that a non-utilitarian, say a philosopher of the type of Sir David Ross, might agree with us in the evaluation of the relative merits of the total sets of consequences of the actions A and B and yet disagree with us about whether we ought to do A or B. He might agree with us in the evaluation of total consequences but disagree with us in the evaluation of possible actions, He might say: “The total consequences of A are better than the total consequences of B, but it would be unjust to do A, for you promised to do B.” My chief concern in this study is with the second type of evaluation: the evaluation of actions. The utilitarian addres- ses himself to people who very likely agree with him as to what consequences are good ones, but who disagree with him about the principle that what we ought to do is to produce the best consequences. For a reason, which will appear presently, the difference between ideal and hedon- istic utilitarianism in most cases will not usually lead to a serious disagreement about what ought to be done in prac- tice. In this section, however, I wish to clear the ground by saying something about the first type of evaluation, the evaluation of consequences. It is with respect to this evalua- tion that Bentham, Mill and Moore differ from one another.3. Hedonistic and non-hedonistic utilitarianism 1s Let us consider Mill’s contention that it is ‘better to be Socrates dissatisfied than a fool satisfied’ Mill holds that pleasure is not to be our sole criterion for evaluating con- sequences: the state of mind of Socrates might be less pleasurable than that of the fool, but, according to Mill, Socrates would be happier than the fool. It is necessary to observe, first of all, that a purely hedon- istic utilitarian, like Bentham, might agree with Mill in preferring the experiences of discontented philosophers to those of contented fools. His preference for the philosopher’s state of mind, however, would not be an intrinsic one. He would say that the discontented philosopher is a useful agent in society and that the existence of Socrates is res- ponsible for an improvement in the lot of humanity gener- ally. Consider two brothers. One may be of a docile and easy temperament: he may lead a supremely contented and unambitious life, enjoying himself hugely, The other brother may be ambitious, may stretch his talents to the full, may strive for scientific success and academic honours, and may discover some invention or some remedy for disease or improvement in agriculture which will enable innumerable men of easy temperament to lead a contented life, whereas otherwise they would have been thwarted by poverty, dis- ease or hunger, Or he may make some advance in pure science which will later have beneficial practical applica- tions. Or, again, he may write poctry which will solace the leisure hours and stimulate the brains of practical men or scientists, thus indirectly leading to an improvement in society. That is, the pleasures of poetry or mathematics may be extrinsically valuable in a way in which those of pushpin or sun-bathing may not be. Though the poet or 1 Utilitarianism, p. 9. The problem of the unhappy sage and the happy fool is cleverly stated in Voltaire’s “Histoire d’un bon Bramin’, Choix de Contes, edited with an introduction and notes by F. C. Green (Cambridge University Press, London, 1951), pp. 245-7.16 An outline of a system of utilitarian ethics mathematician may be discontented, society as a whole may be the more contented for his presence. Again, a man who enjoys pushpin is likely eventually to become bored with it, whereas the man who enjoys poetry is likely to retain this interest throughout his life. Moreover the reading of poetry may develop imagination and sensi- tivity, and so as a result of his interest in poetry a man may be able to do more for the happiness of others than if he had played pushpin and let his brain deteriorate. In short, both for the man immediately concerned. and for others, the pleasures of poetry are, to use Bentham’s word, more fecund than those of pushpin. Perhaps, then, our preference for poetry over pushpin is not one of intrinsic value, but is merely one of extrinsic value. Perhaps strictly in itself and at a particular moment, a contented sheep is as good as a contented philosopher. How- ever it is hard to agree to this. If we did we should have to agree that the human population ought ideally to be reduced by contraceptive methods and the sheep population more than correspondingly increased. Perhaps just so many humans should be left as could keep innumerable millions of placid sheep in contented idleness and immunity from depredations by ferocious animals, Indeed if a contented idiot is as good as a contented philosopher, and if a con- tented sheep is as good as a contented idiot, then a contented fish is as good as a contented sheep, and a contented beetle is as good as a contented fish. Where shall we stop? Maybe we have gone wrong in talking of pleasure as though it were no more than contentment, Contentment consists roughly in relative absence of unsatisfied desires; pleasure is perhaps something more positive and consists in a balance between absence of unsatisfied desires and presence of satisfied desires. We might put the difference in this way: pure unconsciousness would be a limiting case of content- ment, but not of pleasure. A stone has no unsatisfied desires,3. Hedonistic and non-hedonistic utilitarianism 7 but then it just has no desires. Nevertheless, this consideration will not resolve the disagreement between Bentham and. Mill. No doubt a dog has as intense a desire to discover rats as the philosopher has to discover the mysteries of the uni- verse. Mill would wish to say that the pleasures of the philosopher were more valuable intrinsically than those of the dog, however intense these last might be. It appears, then, that many of us may well have a pre~ ference not only for enjoyment as such but for certain sorts of enjoyment. And this goes for many of the humane and beneficent readers whom I am addressing. I suspect that they too have an intrinsic preference for the more complex and intellectual pleasures. This is not surprising. We must not underrate the mere brute strength of a hard and fit human being: by any standards man is a large and strong animal, Nevertheless above all else man owes his survival to his superior intelligence. If man were not a species which was inclined above all else to think and strive, we should not be where we are now. No wonder that men have a liking for intelligence and complexity, and this may become increasingly so in future. Perhaps some people may feel that my remarks here are somewhat too complacent, in view of the liking of so many people for low-grade entertain- ments, such as certain popular television programmes. But even the most avid television addict probably enjoys solving practical problems connected. with his car, his furniture, or his garden. However unintellectual he might be, he would certainly resent the suggestion that he should, if it were possible, change places with a contented sheep, or even a lively and happy dog. Nevertheless, when all is said and done, we must not disguise the fact that disagreements in ultimate attitude are possible between those who like Mill have, and those who like Bentham have not, an intrinsic preference for the ‘higher’ pleasures. However it is possible for two people to disagree about ultimate ends and yet agree18 An outline of a system of utilitarian ethics in practice about what ought to be done. It is worth while enquiring how much practical ethics is likely to be affected by the possibility of disagreement over the question of Socrates dissatisfied versus the fool satisfied. “Not very much’, one feels like saying at first. We noted that the most complex and intellectual pleasures are also the most fecund. Poetry elevates the mind, makes one more sensitive, and so harmonizes with various intellectual pur- suits, some of which are of practical value. Delight in mathe- matics is even more obviously, on Benthamite views, a pleasure worth encouraging, for on the progress of mathe- matics depends the progress of mankind. Even the most hedonistic schoolmaster would prefer to see his boys enjoy- ing poetry and mathematics rather than neglecting these arts for the pleasures of marbles or the tuckshop. Indeed many of the brutish pleasures not only lack fecundity but are actually the reverse of fecund. To enjoy food too much is to end up fat, unhealthy and without zest or vigour, To enjoy drink too much is even worse. In most circumstances of ordinary life the pure hedonist will agree in his practical recommendations with the quasi-ideal utilitarian. This need not always be so. Some years ago two psycho- logists, Olds and Milner, carried out some experiments with rats! Through the skull of each rat they inserted an elec- trode, These electrodes penetrated to various regions of the brain. In the case of some of these regions the rat showed. ‘James Olds and Peter Milner, ‘Positive reinforcement produced by electrical stimulation of the septal area and other regions of the rat brain’, Journal of Comparative and Physiological Psychology 47 (1954) 419-27. James Olds, ‘A preliminary mapping of electrical reinforeing effect in the rat brain’, ibid. 49 (1956) 281-5. L. J. Good has also used. these results of Olds and Milner in order to discuss ethical hedonism. See his ‘A. problem for the hedonist’, in I. J. Good (ed.), ‘The Scientist Speculates (Heinemann London, 1962). Good takes the possibility of this sort of thing to provide a reductio ad absurdum of hedonism.